Language Lab Reproducibility Demo

This tutorial uses data and reproduces a subset of analyses reported in the following manuscript:

Mesquiti & Seraj. (Preprint) The Psychological Impacts of the COVID-19 Pandemic on Corporate Leadership

You can find this project’s github here.

The COVID-19 pandemic sent shockwaves across the fabric of our society. Examining the impact of the pandemic on business leadership is particularly important to understanding how this event affected their decision-making. The present study documents the psychological effects of the COVID-19 pandemic on chief executive officers (CEOs). This was accomplished by analyzing CEOs’ language from quarterly earnings calls (N = 19,536) for a year before and after lockdown. CEOs had large shifts in language in the months immediately following the start of the pandemic lockdowns. Analytic thinking plummeted after the world went into lockdown, with CEOs’ language becoming less technical and more personal and intuitive. In parallel, CEOs’ language showed signs of increased cognitive load, as they were processing the effect of the pandemic on their business practices. Business leaders’ use of collective-focused language (we-usage) dropped substantially after the pandemic began, perhaps suggesting CEOs felt disconnected from their companies. Self-focused (I-usage) language increased, showing the increased preoccupation of business leaders. The size of the observed shifts in language during the pandemic also dwarfed responses to other events that occurred dating back to 2010, with the effect lasting around seven months.

Prep data

Load necessary packages and set Working Directory

if (!require("pacman")) install.packages("pacman")
pacman::p_load(tidyverse,zoo,lubridate,plotrix,ggpubr, caret, broom, kableExtra, reactable, effsize, install = T)
setwd("~/Desktop/Language_Lab_Repro")

Define aesthetics

palette_map = c("#3B9AB2", "#EBCC2A", "#F21A00")
palette_condition = c("#ee9b00", "#bb3e03", "#005f73")

plot_aes = theme_classic() +
  theme(text = element_text(size = 16, family = "Futura Medium")) + 
  theme(axis.text.x=element_text(angle=45, hjust=1)) +
  theme(plot.title.position = 'plot', 
        plot.title = element_text(hjust = 0.5, face = "bold", size = 16)) + 
  theme(axis.text=element_text(size=16),
        axis.title=element_text(size=20,face="bold"))+
  theme(plot.title.position = 'plot', 
        plot.title = element_text(hjust = 0.5, face = "bold", size = 20)) +
  theme(axis.text=element_text(size = 14),
        axis.title=element_text(size = 20,face="bold"))

Write our Table Funcions

baseline_ttest <- function(ttest_list) {
  # Extract relevant information from each test and store in a data frame
  ttest_df <- data.frame(
    Group1 = seq(0,0,1),
    Group2 = seq(1,24,1),
    t = sapply(ttest_list, function(x) x$statistic),
    df = sapply(ttest_list, function(x) x$parameter),
    p_value = sapply(ttest_list, function(x) x$p.value)
  )
  
  # Format p-values as scientific notation
  ttest_df$p_value <- format(ttest_df$p_value, scientific = T)
  
  # Rename columns
  colnames(ttest_df) <- c("t", "t + 1 ", "t-value", "Degrees of Freedom", "p-value")
  
  # Create table using kableExtra
  kable(ttest_df, caption = "Summary of Welch's t-Tests", booktabs = TRUE) %>%
   kableExtra::kable_styling()
}

post_pandemic_summary <- function(ttest_list) {
  # Extract relevant information from each test and store in a data frame
  ttest_df <- data.frame(
    Group1 = seq(12,23,1),
    Group2 = seq(13,24,1),
    t = sapply(ttest_list, function(x) x$statistic),
    df = sapply(ttest_list, function(x) x$parameter),
    p_value = sapply(ttest_list, function(x) x$p.value)
  )
  
  # Format p-values as scientific notation
  ttest_df$p_value <- format(ttest_df$p_value, scientific = T)
  
  # Rename columns
  colnames(ttest_df) <- c("t", "t + 1 ", "t-value", "Degrees of Freedom", "p-value")
  
  # Create table using kableExtra
  kable(ttest_df, caption = "Summary of Welch's t-Tests", booktabs = TRUE) %>%
   kableExtra::kable_styling()
}



baseline_cohen_d <- function(cohen_d_list) {
  # Extract relevant information from each test and store in a data frame
  cohen_d_df <- data.frame(
    Group1 = seq(0,0,1),
    Group2 = seq(1,24,1),
    Cohen_d = sapply(cohen_d_list, function(x) x$estimate)
  )
  
  # Rename columns
  colnames(cohen_d_df) <- c("t", "t + 1", "Cohen's d")
  
  # Create table using kableExtra
  kable(cohen_d_df, caption = "Summary of Cohen's D", booktabs = TRUE) %>%
   kableExtra::kable_styling()
}

post_cohen_d <- function(cohen_d_list) {
  # Extract relevant information from each test and store in a data frame
  cohen_d_df <- data.frame(
    Group1 = seq(12,23,1),
    Group2 = seq(13,24,1),
    Cohen_d = sapply(cohen_d_list, function(x) x$estimate)
  )
  
  # Rename columns
  colnames(cohen_d_df) <- c("t", "t+1", "Cohen's d")
  
  # Create table using kableExtra
  kable(cohen_d_df, caption = "Summary of Cohen's D", booktabs = TRUE) %>%
   kableExtra::kable_styling()
}

baseline_mean_diff <- function(mean_diff_list) {
  # Extract relevant information from each mean difference calculation and store in a data frame
  mean_diff_df <- data.frame(
    Group1 = seq(0,0,1),
    Group2 = seq(1,24,1),
    mean_diff = mean_diff_list
  )
  
  # Rename columns
  colnames(mean_diff_df) <- c("t", "t+1", "Mean Difference")
  
  # Create table using kableExtra
  kable(mean_diff_df, caption = "Summary of Mean Differences", booktabs = TRUE) %>%
   kableExtra::kable_styling()
}


post_mean_diff <- function(mean_diff_list) {
  # Extract relevant information from each mean difference calculation and store in a data frame
  mean_diff_df <- data.frame(
    Group1 = seq(12,23,1),
    Group2 = seq(13,24,1),
    mean_diff = mean_diff_list
  )
  
  # Rename columns
  colnames(mean_diff_df) <- c("t", "t+1", "Mean Difference")
  
  # Create table using kableExtra
  kable(mean_diff_df, caption = "Summary of Mean Differences", booktabs = TRUE) %>%
   kableExtra::kable_styling()
}

Load in the Data

data  <-  read_csv("https://raw.githubusercontent.com/scm1210/Language_Lab_Repro/main/Big_CEO.csv") #read in the data from github 

data <- data["2019-03-01"<= data$Date & data$Date <= "2021-04-01",] #subsetting covid dates 

data <- data %>% filter(WC<=5400) %>% #filter out based on our exclusion criteria
  filter(WC>=25)

data$month_year <- format(as.Date(data$Date), "%Y-%m") #reformat 

data_tidy <- data %>% dplyr::select(Date, Speaker, Analytic, cogproc,allnone,we,i,emo_anx) %>%
  mutate(Date = lubridate::ymd(Date),
         time_month = as.numeric(Date - ymd("2019-03-01")) / 30, #centering at start of march
         time_month_quad = time_month * time_month) #making our quadratic term

data_tidy$Date_off <- floor(data_tidy$time_month) #rounding off dates to whole months using ceiling function (0 = 2019-03, 24 = 2021-04)
data_tidy$Date_covid <- as.factor(data_tidy$Date_off) #factorize

Create Tidy Data for Graphs

df <- read_csv("https://raw.githubusercontent.com/scm1210/Language_Lab_Repro/main/Big_CEO.csv")#put code here to read in Big CEO data
df <- df %>% filter(WC<=5400)   %>% 
  filter(WC>=25)

df$month_year <- format(as.Date(df$Date), "%Y-%m") ###extracting month and year to build fiscal quarter graphs, need a new variable bc if not it'll give us issues

df2 <- df %>%#converting our dates to quarterly dates 
  group_by(month_year) %>% ###grouping by the Top100 tag and date 
  summarise_at(vars("Date","WC","Analytic","cogproc",'we','i'),  funs(mean, std.error),) #pulling the means and SEs for our variables of interest

df2 <- df2["2019-01"<= df2$month_year & df2$month_year <= "2021-03",] #covid dates 

Write our Stats Functions

We were interested in how language changed relative to baseline one year pre-pandemic, as well as how language changed after the Pandemic.

As a result we ran two separate set of analyses comparing t(time zero) to t[i] and t(12 months after our centered data point) to t + 1. The groups you see will be centered on 03/2019. That is, 12 = 03/2020, 13 = 04/2020, etc. etc.

Analytic Thinking

analytic_my.t = function(fac1, fac2){
  t.test(data_tidy$Analytic[data_tidy$Date_covid==fac1], 
         data_tidy$Analytic[data_tidy$Date_covid==fac2])
} #writing our t-test function to compare t to t[i] 

analytic_my.d = function(fac1, fac2){
  cohen.d(data_tidy$Analytic[data_tidy$Date_covid==fac1], 
          data_tidy$Analytic[data_tidy$Date_covid==fac2])
} #function for cohen's d

analytic_mean <-  function(fac1, fac2){
  mean(data_tidy$Analytic[data_tidy$Date_covid==fac1])- 
    mean(data_tidy$Analytic[data_tidy$Date_covid==fac2])
} #function to do mean differences

Cognitive Processing

cogproc_my.t = function(fac1, fac2){
  t.test(data_tidy$cogproc[data_tidy$Date_covid==fac1], 
         data_tidy$cogproc[data_tidy$Date_covid==fac2])
} #writing our t-test function to compare t to t[i] 


cogproc_my.d = function(fac1, fac2){
  cohen.d(data_tidy$cogproc[data_tidy$Date_covid==fac1], 
          data_tidy$cogproc[data_tidy$Date_covid==fac2])
} #function for cohen's d

cogproc_mean <-  function(fac1, fac2){
  mean(data_tidy$cogproc[data_tidy$Date_covid==fac1])- 
    mean(data_tidy$cogproc[data_tidy$Date_covid==fac2])
} #function to do mean differences

I-words

i_my.t = function(fac1, fac2){
  t.test(data_tidy$i[data_tidy$Date_covid==fac1], 
         data_tidy$i[data_tidy$Date_covid==fac2])
} #writing our t-test function to compare t to t + 1 

i_my.d = function(fac1, fac2){
  cohen.d(data_tidy$i[data_tidy$Date_covid==fac1], 
          data_tidy$i[data_tidy$Date_covid==fac2])
} #function for cohen's d


i_mean <-  function(fac1, fac2){
  mean(data_tidy$i[data_tidy$Date_covid==fac1])- 
    mean(data_tidy$i[data_tidy$Date_covid==fac2])
} #function to do mean differences

We-words

we_my.t = function(fac1, fac2){
  t.test(data_tidy$we[data_tidy$Date_covid==fac1], 
         data_tidy$we[data_tidy$Date_covid==fac2])
} 

we_my.d = function(fac1, fac2){
  cohen.d(data_tidy$we[data_tidy$Date_covid==fac1], 
          data_tidy$we[data_tidy$Date_covid==fac2])
} #function for cohen's d

we_mean <-  function(fac1, fac2){
  mean(data_tidy$we[data_tidy$Date_covid==fac1])- 
    mean(data_tidy$we[data_tidy$Date_covid==fac2])
} #function to do mean differences

Tidy data

Data transformations

• None

Exclusions

• Excluded texts that were shorter than ** 25 words ** and greater than ** 5,400 words **!

Summary of the Data

Range of Dates

range(data$Date)

## [1] "2019-03-01" "2021-04-01"

Number of Speakers

speakers <- data %>%
  select(Speaker) %>%
  unique() %>%
  dplyr::summarize(n = n()) %>%
  reactable::reactable(striped = TRUE)
speakers

Number of Transcripts

transcripts <- data %>%
  select(1) %>%
  dplyr::summarize(n = n()) %>%
  reactable::reactable(striped = TRUE)
transcripts

Mean Word Count

word_count <- data %>%
  select(WC) %>%
  dplyr::summarize(mean = mean(WC)) %>%
  reactable::reactable(striped = TRUE)
word_count

How did language change after the Pandemic?

Analytic Thinking

T-test

analytic_ttest<- mapply(analytic_my.t,seq(12,23,1), seq(13,24,1),SIMPLIFY=F) #compare t (first parantheses) to t[i] (second parentheses)increasing by 1
post_pandemic_summary(analytic_ttest)

Summary of Welch’s t-Tests

t	t + 1	t-value	Degrees of Freedom	p-value
12	13	5.0848767	525.79320	5.124345e-07
13	14	-2.5947675	373.06404	9.838752e-03
14	15	-1.6725600	252.03534	9.565479e-02
15	16	1.9241684	377.61680	5.508471e-02
16	17	-2.2121608	200.57005	2.808412e-02
17	18	-1.6872236	218.93267	9.298455e-02
18	19	0.6199376	262.60906	5.358364e-01
19	20	0.8737898	128.21711	3.838664e-01
20	21	-1.5397962	230.75591	1.249802e-01
21	22	1.9533418	94.31745	5.374259e-02
22	23	-1.1497811	55.55164	2.551600e-01
23	24	-1.7179003	2141.37219	8.595937e-02

Cohen’s D

analytic_d <- mapply(analytic_my.d,seq(12,23,1), seq(13,24,1),SIMPLIFY=FALSE) 
post_cohen_d(analytic_d)

Summary of Cohen’s D

t	t+1	Cohen’s d
12	13	0.3274589
13	14	-0.1597933
14	15	-0.1320224
15	16	0.1935631
16	17	-0.1616992
17	18	-0.1481301
18	19	0.0709701
19	20	0.0898748
20	21	-0.1246402
21	22	0.2681803
22	23	-0.1598304
23	24	-0.0739462

Mean Differences

analytic_meandiff <- mapply(analytic_mean, seq(12,23,1), seq(13,24,1)) #across all of the months comparing to time zero
post_mean_diff(analytic_meandiff)

Summary of Mean Differences

t	t+1	Mean Difference
12	13	4.734622
13	14	-2.190455
14	15	-1.844328
15	16	2.748318
16	17	-2.231753
17	18	-2.101267
18	19	1.158869
19	20	1.276462
20	21	-1.779122
21	22	4.065080
22	23	-2.075629
23	24	-0.994088

Cogproc

T-test

cogproc_ttest <-mapply(cogproc_my.t, seq(12,23,1), seq(13,24,1),SIMPLIFY=FALSE) #compare t (first parathese) to t[i] (second parantheses) increasing by 1
post_pandemic_summary(cogproc_ttest)

Summary of Welch’s t-Tests

t	t + 1	t-value	Degrees of Freedom	p-value
12	13	-4.3160945	534.57336	1.892660e-05
13	14	1.4046015	366.53625	1.609866e-01
14	15	4.0193476	257.86515	7.665356e-05
15	16	-3.1317117	367.29975	1.877275e-03
16	17	0.9867920	199.23919	3.249415e-01
17	18	4.1803820	223.61017	4.177506e-05
18	19	-1.1984064	285.88282	2.317513e-01
19	20	-1.4929615	133.61894	1.378047e-01
20	21	3.2109343	234.84605	1.508000e-03
21	22	-1.7045407	87.34608	9.183489e-02
22	23	0.9967763	55.37573	3.232089e-01
23	24	-0.9994281	2145.12748	3.177001e-01

Cohen’s D

cogproc_d <-mapply(cogproc_my.d, seq(12,23,1), seq(13,24,1),SIMPLIFY=FALSE)
post_cohen_d(cogproc_d)

Summary of Cohen’s D

t	t+1	Cohen’s d
12	13	-0.2755415
13	14	0.0887056
14	15	0.3007241
15	16	-0.3204553
16	17	0.0732556
17	18	0.3435609
18	19	-0.1329353
19	20	-0.1294167
20	21	0.2476709
21	22	-0.2453381
22	23	0.1405453
23	24	-0.0429758

Mean Differences

cogproc_meandiff <- mapply(cogproc_mean, seq(12,23,1), seq(13,24,1)) # comparing time zero [3/2019]across all of the months
post_mean_diff(cogproc_meandiff)

Summary of Mean Differences

t	t+1	Mean Difference
12	13	-0.6107287
13	14	0.1784774
14	15	0.6094504
15	16	-0.6540232
16	17	0.1559844
17	18	0.7442075
18	19	-0.2962170
19	20	-0.2746360
20	21	0.5304979
21	22	-0.5357971
22	23	0.2775877
23	24	-0.0886600

I-words

T-test

i_ttest <- mapply(i_my.t, seq(12,23,1), seq(13,24,1),SIMPLIFY=FALSE) #compare t (first paratheses) to t[i] (second parentheses) increasing by 1
post_pandemic_summary(i_ttest)

Summary of Welch’s t-Tests

t	t + 1	t-value	Degrees of Freedom	p-value
12	13	-5.1026305	477.85082	4.841738e-07
13	14	2.9682570	362.96961	3.193717e-03
14	15	2.7352278	261.20479	6.660709e-03
15	16	-3.5894844	336.98113	3.805206e-04
16	17	1.7614255	191.52014	7.976208e-02
17	18	3.4393905	240.73312	6.870032e-04
18	19	-2.6019091	255.11065	9.812584e-03
19	20	0.4503223	134.90596	6.532009e-01
20	21	1.5059378	248.77332	1.333518e-01
21	22	2.0158644	84.28386	4.699962e-02
22	23	-3.8068297	57.55886	3.436805e-04
23	24	4.4094793	2135.84040	1.087616e-05

Cohen’s D

i_d <- mapply(i_my.d,seq(12,23,1), seq(13,24,1),SIMPLIFY=FALSE)
post_cohen_d(i_d)

Summary of Cohen’s D

t	t+1	Cohen’s d
12	13	-0.3467518
13	14	0.1902125
14	15	0.1990807
15	16	-0.3757604
16	17	0.1451672
17	18	0.2369631
18	19	-0.3007221
19	20	0.0377993
20	21	0.1020099
21	22	0.2971566
22	23	-0.4621942
23	24	0.1900173

Mean Differences

i_meandiff <- mapply(i_mean,seq(12,23,1), seq(13,24,1)) # comparing time zero [3/2020]across all of the months
post_mean_diff(i_meandiff)

Summary of Mean Differences

t	t+1	Mean Difference
12	13	-0.2878044
13	14	0.1550533
14	15	0.1624754
15	16	-0.3241516
16	17	0.1289192
17	18	0.2083141
18	19	-0.2363725
19	20	0.0329017
20	21	0.0885966
21	22	0.2292627
22	23	-0.3911951
23	24	0.1657095

We-words

T-test

we_ttest <- mapply(we_my.t, seq(12,23,1), seq(13,24,1),SIMPLIFY=FALSE) #compare t (first parathese) to t[i] (second parantheses) increasing by 1
post_pandemic_summary(we_ttest)

Summary of Welch’s t-Tests

t	t + 1	t-value	Degrees of Freedom	p-value
12	13	4.1037791	527.07583	4.708824e-05
13	14	0.9116989	378.81928	3.625070e-01
14	15	-3.3226285	253.13940	1.023448e-03
15	16	2.4647106	373.96103	1.416113e-02
16	17	-0.3375119	197.51750	7.360894e-01
17	18	-4.2758502	229.49548	2.793946e-05
18	19	2.5509775	262.60210	1.130991e-02
19	20	-0.1421962	131.79434	8.871422e-01
20	21	-1.9395335	238.21223	5.361708e-02
21	22	-0.2952385	84.06212	7.685396e-01
22	23	0.8556597	55.76358	3.958478e-01
23	24	-0.3495394	2137.76534	7.267188e-01

Cohen’s D

we_d <- mapply(we_my.d, seq(12,23,1), seq(13,24,1),SIMPLIFY=FALSE)
post_cohen_d(we_d)

Summary of Cohen’s D

t	t+1	Cohen’s d
12	13	0.2639367
13	14	0.0549934
14	15	-0.2594704
15	16	0.2501259
16	17	-0.0255875
17	18	-0.3276203
18	19	0.2920369
19	20	-0.0129636
20	21	-0.1443587
21	22	-0.0435999
22	23	0.1169953
23	24	-0.0150573

Mean Differences

we_meandiff <- mapply(we_mean, seq(12,23,1), seq(13,24,1)) # comparing time zero [3/2020]across all of the months
post_mean_diff(we_meandiff)

Summary of Mean Differences

t	t+1	Mean Difference
12	13	0.3777932
13	14	0.0763380
14	15	-0.3676046
15	16	0.3649285
16	17	-0.0365235
17	18	-0.4710551
18	19	0.4168557
19	20	-0.0182846
20	21	-0.2041654
21	22	-0.0608833
22	23	0.1582888
23	24	-0.0209555

How did language change relative to baseline (one year before the pandemic; 03/2019)?

Analytic Thining

T-test

analytic_ttest_baseline <-mapply(analytic_my.t,0, seq(1,24,1),SIMPLIFY=FALSE) #compare t (first parantheses) to t[i] (second parentheses)increasing by 1
baseline_ttest(analytic_ttest_baseline)

Summary of Welch’s t-Tests

t	t + 1	t-value	Degrees of Freedom	p-value
0	1	1.5025175	1161.46333	1.332353e-01
0	2	0.6860139	1036.84856	4.928577e-01
0	3	0.2507930	245.14330	8.021842e-01
0	4	2.6728372	1120.10414	7.630544e-03
0	5	0.4785480	1004.80108	6.323643e-01
0	6	1.0343183	280.42507	3.018785e-01
0	7	2.6674817	1049.94370	7.759826e-03
0	8	1.4045584	993.35140	1.604652e-01
0	9	1.0147460	328.09331	3.109746e-01
0	10	1.5505263	286.24028	1.221201e-01
0	11	1.9737798	1061.63898	4.866575e-02
0	12	1.3053906	1272.10102	1.919959e-01
0	13	5.7769548	623.93692	1.200948e-08
0	14	5.1516350	929.47739	3.153290e-07
0	15	1.4218984	370.16499	1.558977e-01
0	16	3.9258380	316.92397	1.060657e-04
0	17	3.2571976	918.08556	1.166437e-03
0	18	0.1171218	302.23369	9.068413e-01
0	19	0.8462858	164.42299	3.986233e-01
0	20	3.7364297	920.43945	1.981471e-04
0	21	0.6393118	331.79337	5.230612e-01
0	22	2.6168435	63.20064	1.108971e-02
0	23	3.7686675	1111.95144	1.727388e-04
0	24	2.4325523	1125.18803	1.514789e-02

Cohen’s D

analytic_D_baseline <- mapply(analytic_my.d,0, seq(1,24,1),SIMPLIFY=FALSE) 
baseline_cohen_d(analytic_D_baseline)

Summary of Cohen’s D

t	t + 1	Cohen’s d
0	1	0.0879752
0	2	0.0329980
0	3	0.0206107
0	4	0.1587215
0	5	0.0235235
0	6	0.0867045
0	7	0.1620807
0	8	0.0687147
0	9	0.0805849
0	10	0.1282654
0	11	0.1023933
0	12	0.0694416
0	13	0.3954264
0	14	0.2534133
0	15	0.1138341
0	16	0.3057368
0	17	0.1588173
0	18	0.0101558
0	19	0.0861013
0	20	0.1802980
0	21	0.0529819
0	22	0.3237240
0	23	0.2018620
0	24	0.1262979

Mean Differences

analytic_mean_baseline <- mapply(analytic_mean, 0, seq(1,24,1)) #across all of the months comparing to time zero
baseline_mean_diff(analytic_mean_baseline)

Summary of Mean Differences

t	t+1	Mean Difference
0	1	1.3114081
0	2	0.4935284
0	3	0.3039970
0	4	2.3251490
0	5	0.3411544
0	6	1.3027809
0	7	2.3954214
0	8	0.9976299
0	9	1.1986758
0	10	1.9188652
0	11	1.4369448
0	12	1.0438407
0	13	5.7784625
0	14	3.5880071
0	15	1.7436794
0	16	4.4919977
0	17	2.2602447
0	18	0.1589776
0	19	1.3178462
0	20	2.5943085
0	21	0.8151869
0	22	4.8802673
0	23	2.8046380
0	24	1.8105501

Cogproc

T-test

cogproc_ttest_baseline <- mapply(cogproc_my.t, 0, seq(1,24,1),SIMPLIFY=FALSE) #compare t (first parathese) to t[i] (second parantheses) increasing by 1
baseline_ttest(cogproc_ttest_baseline)

Summary of Welch’s t-Tests

t	t + 1	t-value	Degrees of Freedom	p-value
0	1	-0.5097155	1156.50973	6.103480e-01
0	2	-0.7178587	1035.96962	4.730063e-01
0	3	-0.2391309	218.72044	8.112280e-01
0	4	-1.8416817	1119.69687	6.578607e-02
0	5	-0.3763500	1051.93803	7.067326e-01
0	6	0.2442296	282.79380	8.072301e-01
0	7	-1.7141683	1029.21251	8.679890e-02
0	8	-0.9538148	1076.64206	3.403915e-01
0	9	1.0445702	320.30692	2.970093e-01
0	10	-0.8168779	255.25892	4.147599e-01
0	11	-0.7245359	1147.57474	4.688845e-01
0	12	-2.0279981	1307.90475	4.276280e-02
0	13	-5.7012479	609.24510	1.854777e-08
0	14	-6.5910797	924.04251	7.328808e-11
0	15	-0.3855580	395.99482	7.000311e-01
0	16	-4.0811802	298.22073	5.758392e-05
0	17	-5.4650159	949.00294	5.916345e-08
0	18	0.9264753	310.66778	3.549182e-01
0	19	-0.5797074	184.73797	5.628182e-01
0	20	-3.7993649	936.80835	1.544264e-04
0	21	0.7639021	341.61474	4.454529e-01
0	22	-1.3820415	61.97292	1.719203e-01
0	23	-1.0690564	1140.02341	2.852706e-01
0	24	-1.8593187	1172.33479	6.323237e-02

Cohen’s D

cogproc_D_baseline <- mapply(cogproc_my.d, 0, seq(1,24,1),SIMPLIFY=FALSE)
baseline_cohen_d(cogproc_D_baseline)

Summary of Cohen’s D

t	t + 1	Cohen’s d
0	1	-0.0298959
0	2	-0.0345459
0	3	-0.0213194
0	4	-0.1093919
0	5	-0.0180369
0	6	0.0203613
0	7	-0.1048291
0	8	-0.0445936
0	9	0.0841121
0	10	-0.0731906
0	11	-0.0364241
0	12	-0.1070381
0	13	-0.3938811
0	14	-0.3255788
0	15	-0.0297828
0	16	-0.3291694
0	17	-0.2601030
0	18	0.0788773
0	19	-0.0527050
0	20	-0.1809343
0	21	0.0622160
0	22	-0.1777619
0	23	-0.0568265
0	24	-0.0951265

Mean Differences

cogproc_mean_baseline <- mapply(cogproc_mean, 0, seq(1,24,1)) # comparing time zero [3/2020]across all of the months
baseline_mean_diff(cogproc_meandiff)

Summary of Mean Differences

t	t+1	Mean Difference
0	1	-0.6107287
0	2	0.1784774
0	3	0.6094504
0	4	-0.6540232
0	5	0.1559844
0	6	0.7442075
0	7	-0.2962170
0	8	-0.2746360
0	9	0.5304979
0	10	-0.5357971
0	11	0.2775877
0	12	-0.0886600
0	13	-0.6107287
0	14	0.1784774
0	15	0.6094504
0	16	-0.6540232
0	17	0.1559844
0	18	0.7442075
0	19	-0.2962170
0	20	-0.2746360
0	21	0.5304979
0	22	-0.5357971
0	23	0.2775877
0	24	-0.0886600

I-words

T-test

i_ttest_baseline <- mapply(i_my.t, 0, seq(1,24,1),SIMPLIFY=FALSE) #compare t (first paratheseses) to t[i] (second parentheses) increasing by 1
baseline_ttest(i_ttest_baseline)

Summary of Welch’s t-Tests

t	t + 1	t-value	Degrees of Freedom	p-value
0	1	-3.3449936	1143.81760	8.495412e-04
0	2	-1.1963077	1155.18280	2.318220e-01
0	3	-0.1911368	213.55326	8.486000e-01
0	4	-4.1439455	1114.30669	3.672274e-05
0	5	-0.6476795	1056.55931	5.173329e-01
0	6	-1.6111266	278.02962	1.082868e-01
0	7	-3.3533234	1035.23122	8.273950e-04
0	8	-2.0582130	1066.95830	3.981213e-02
0	9	-1.4167584	265.19170	1.577272e-01
0	10	-2.7747321	284.30487	5.890772e-03
0	11	-1.9848824	1154.30486	4.739397e-02
0	12	-0.3319999	1263.49829	7.399444e-01
0	13	-5.0279656	571.48514	6.644118e-07
0	14	-3.7092732	958.88047	2.197939e-04
0	15	0.2214347	390.57770	8.248697e-01
0	16	-3.9254650	253.43509	1.115955e-04
0	17	-4.4733002	1005.42169	8.580050e-06
0	18	0.4134953	350.62439	6.794966e-01
0	19	-2.6459757	180.59824	8.864330e-03
0	20	-4.3779164	986.11105	1.326045e-05
0	21	-1.3221651	371.12983	1.869275e-01
0	22	1.3507586	63.33638	1.815790e-01
0	23	-5.6222563	1250.83820	2.322252e-08
0	24	-1.8930601	1254.79690	5.857980e-02

Cohen’s D

i_D_baseline <- mapply(i_my.d, 0, seq(1,24,1),SIMPLIFY=FALSE)
baseline_cohen_d(i_D_baseline)

Summary of Cohen’s D

t	t + 1	Cohen’s d
0	1	-0.1965974
0	2	-0.0543981
0	3	-0.0173720
0	4	-0.2467407
0	5	-0.0309676
0	6	-0.1358241
0	7	-0.2047181
0	8	-0.0966976
0	9	-0.1296303
0	10	-0.2305339
0	11	-0.0995545
0	12	-0.0176937
0	13	-0.3562055
0	14	-0.1785725
0	15	0.0172266
0	16	-0.3536629
0	17	-0.2047237
0	18	0.0327380
0	19	-0.2453415
0	20	-0.2010721
0	21	-0.1028381
0	22	0.1664219
0	23	-0.2903836
0	24	-0.0945412

Mean Differences

i_mean_baseline <- mapply(i_mean, 0, seq(1,24,1)) # comparing time zero [3/2020]across all of the months
baseline_mean_diff(i_mean_baseline)

Summary of Mean Differences

t	t+1	Mean Difference
0	1	-0.1747670
0	2	-0.0504304
0	3	-0.0148774
0	4	-0.2082233
0	5	-0.0265697
0	6	-0.1159251
0	7	-0.1744079
0	8	-0.0846426
0	9	-0.1162156
0	10	-0.1958046
0	11	-0.0842683
0	12	-0.0149918
0	13	-0.3027962
0	14	-0.1477429
0	15	0.0147325
0	16	-0.3094191
0	17	-0.1804999
0	18	0.0278142
0	19	-0.2085583
0	20	-0.1756567
0	21	-0.0870600
0	22	0.1422027
0	23	-0.2489924
0	24	-0.0832828

We-words

T-test

we_ttest_baseline <- mapply(we_my.t, 0, seq(1,24,1),SIMPLIFY=FALSE) #compare t (first parathese) to t[i] (second parantheses) increasing by 1
baseline_ttest(we_ttest_baseline)

Summary of Welch’s t-Tests

t	t + 1	t-value	Degrees of Freedom	p-value
0	1	0.5717846	1161.88366	5.675785e-01
0	2	1.5919359	1008.44599	1.117125e-01
0	3	-1.0685461	214.74566	2.864739e-01
0	4	0.6153736	1116.22594	5.384335e-01
0	5	0.9396382	979.10341	3.476349e-01
0	6	-1.1795694	280.31623	2.391716e-01
0	7	-0.2036380	1067.87587	8.386752e-01
0	8	0.6497173	972.54306	5.160283e-01
0	9	-0.6307460	351.28994	5.286168e-01
0	10	-0.9676900	309.04342	3.339559e-01
0	11	-0.9267542	1073.79066	3.542624e-01
0	12	-0.4001689	1197.17272	6.891035e-01
0	13	3.3603937	676.58875	8.220450e-04
0	14	5.6600766	890.33555	2.040178e-08
0	15	0.4231819	395.82346	6.723924e-01
0	16	3.3898118	317.82041	7.875779e-04
0	17	5.1355858	889.19740	3.456716e-07
0	18	-0.7164395	361.98379	4.741820e-01
0	19	2.3093761	191.37725	2.199015e-02
0	20	4.1802245	873.54302	3.205482e-05
0	21	0.8666887	390.06057	3.866454e-01
0	22	0.2287513	64.77158	8.197829e-01
0	23	2.5427088	1081.13071	1.113820e-02
0	24	2.2872647	1080.95421	2.237292e-02

Cohen’s D

we_D_baseline <- mapply(we_my.d, 0, seq(1,24,1),SIMPLIFY=FALSE)
baseline_cohen_d(we_D_baseline)

Summary of Cohen’s D

t	t + 1	Cohen’s d
0	1	0.0334412
0	2	0.0777773
0	3	-0.0966754
0	4	0.0362120
0	5	0.0468851
0	6	-0.0989057
0	7	-0.0122764
0	8	0.0321927
0	9	-0.0482579
0	10	-0.0764371
0	11	-0.0478523
0	12	-0.0216259
0	13	0.2228626
0	14	0.2873740
0	15	0.0326963
0	16	0.2635803
0	17	0.2566654
0	18	-0.0557482
0	19	0.2039772
0	20	0.2102911
0	21	0.0657068
0	22	0.0270689
0	23	0.1373736
0	24	0.1204946

Mean Differences

we_mean_baseline <- mapply(we_mean, 0, seq(1,24,1)) # comparing time zero [3/2020]across all of the months
baseline_mean_diff(we_mean_baseline)

Summary of Mean Differences

t	t+1	Mean Difference
0	1	0.0530735
0	2	0.1226640
0	3	-0.1575023
0	4	0.0544833
0	5	0.0717923
0	6	-0.1604908
0	7	-0.0190853
0	8	0.0495303
0	9	-0.0765531
0	10	-0.1217559
0	11	-0.0731274
0	12	-0.0334520
0	13	0.3443412
0	14	0.4206792
0	15	0.0530747
0	16	0.4180032
0	17	0.3814797
0	18	-0.0895754
0	19	0.3272803
0	20	0.3089956
0	21	0.1048303
0	22	0.0439469
0	23	0.2022358
0	24	0.1812803

Build our Graphs

Analytic Thinking

Analytic <- ggplot(data=df2, aes(x=Date_mean, y=Analytic_mean, group=1)) +
  geom_line(colour = "dodgerblue3") +
  scale_x_date(date_breaks = "3 month", date_labels = "%Y-%m") +
  geom_ribbon(aes(ymin=Analytic_mean-Analytic_std.error, ymax=Analytic_mean+Analytic_std.error), alpha=0.2) +
  ggtitle("Analytic Thinking") +
  labs(x = "Month", y = 'Standardized score') +
  plot_aes + #here's our plot aes object
  geom_vline(xintercept = as.numeric(as.Date("2020-03-01")), linetype = 1) +
  geom_rect(data = df2, #summer surge
            aes(xmin = as.Date("2020-06-15", "%Y-%m-%d"), 
                xmax = as.Date("2020-07-20",  "%Y-%m-%d"),
                ymin = -Inf, 
                ymax = Inf),
            fill = "gray", 
            alpha = 0.009) +
  geom_rect(data = df2, #winter surge
            aes(xmin = as.Date("2020-11-15", "%Y-%m-%d"), 
                xmax = as.Date("2021-01-01",  "%Y-%m-%d"),
                ymin = -Inf, 
                ymax = Inf),
            fill = "gray", 
            alpha = 0.009)
Analytic <- Analytic + annotate(geom="text",x=as.Date("2020-07-01"),
                                y=43,label="Summer 2020 surge", size = 3) + 
  annotate(geom="text",x=as.Date("2020-12-03"),
           y=43,label="Winter 2020 surge", size = 3)
Analytic

Cogproc

Cogproc <- ggplot(data=df2, aes(x=Date_mean, y=cogproc_mean, group=1)) +
  geom_line(colour = "dodgerblue3") +
  scale_x_date(date_breaks = "3 month", date_labels = "%Y-%m") +
  geom_ribbon(aes(ymin=cogproc_mean-cogproc_std.error, ymax=cogproc_mean+cogproc_std.error), alpha=0.2) +
  ggtitle("Cognitive Processing") +
  labs(x = "Month", y = '% Total Words') +
  plot_aes + #here's our plot aes object
  geom_vline(xintercept = as.numeric(as.Date("2020-03-01")), linetype = 1) +
  geom_rect(data = df2, #summer surge
            aes(xmin = as.Date("2020-06-15", "%Y-%m-%d"), 
                xmax = as.Date("2020-07-20",  "%Y-%m-%d"),
                ymin = -Inf, 
                ymax = Inf),
            fill = "gray", 
            alpha = 0.009) +
  geom_rect(data = df2, #winter surge
            aes(xmin = as.Date("2020-11-15", "%Y-%m-%d"), 
                xmax = as.Date("2021-01-01",  "%Y-%m-%d"),
                ymin = -Inf, 
                ymax = Inf),
            fill = "gray", 
            alpha = 0.009)
Cogproc <- Cogproc + annotate(geom="text",x=as.Date("2020-07-01"),
                                y=12.5,label="Summer 2020 surge", size = 3) + 
  annotate(geom="text",x=as.Date("2020-12-03"),
           y=12.5,label="Winter 2020 surge", size = 3)
Cogproc

I-words

i <- ggplot(data=df2, aes(x=Date_mean, y=i_mean, group=1)) +
  geom_line(colour = "dodgerblue3") +
  scale_x_date(date_breaks = "3 month", date_labels = "%Y-%m") +
  geom_ribbon(aes(ymin=i_mean-i_std.error, ymax=i_mean+i_std.error), alpha=0.2) +
  ggtitle("I-usage") +
  labs(x = "Month", y = '% Total Words') +
  plot_aes + #here's our plot aes object
  geom_vline(xintercept = as.numeric(as.Date("2020-03-01")), linetype = 1) +
  geom_rect(data = df2, #summer surge
            aes(xmin = as.Date("2020-06-15", "%Y-%m-%d"), 
                xmax = as.Date("2020-07-20",  "%Y-%m-%d"),
                ymin = -Inf, 
                ymax = Inf),
            fill = "gray", 
            alpha = 0.009) +
  geom_rect(data = df2, #winter surge
            aes(xmin = as.Date("2020-11-15", "%Y-%m-%d"), 
                xmax = as.Date("2021-01-01",  "%Y-%m-%d"),
                ymin = -Inf, 
                ymax = Inf),
            fill = "gray", 
            alpha = 0.009)
i <- i + annotate(geom="text",x=as.Date("2020-07-01"),
                                y=1.95,label="Summer 2020 surge", size = 3) + 
  annotate(geom="text",x=as.Date("2020-12-03"),
           y=1.95,label="Winter 2020 surge", size = 3)
i

We-words

we <- ggplot(data=df2, aes(x=Date_mean, y=we_mean, group=1)) +
  geom_line(colour = "dodgerblue3") +
  scale_x_date(date_breaks = "3 month", date_labels = "%Y-%m") +
  geom_ribbon(aes(ymin=we_mean-we_std.error, ymax=we_mean+we_std.error), alpha=0.2) +
  ggtitle("We-usage") +
  labs(x = "Month", y = '% Total Words') +
  plot_aes + #here's our plot aes object
  geom_vline(xintercept = as.numeric(as.Date("2020-03-01")), linetype = 1) +
  geom_rect(data = df2, #summer surge
            aes(xmin = as.Date("2020-06-15", "%Y-%m-%d"), 
                xmax = as.Date("2020-07-20",  "%Y-%m-%d"),
                ymin = -Inf, 
                ymax = Inf),
            fill = "gray", 
            alpha = 0.009) +
  geom_rect(data = df2, #winter surge
            aes(xmin = as.Date("2020-11-15", "%Y-%m-%d"), 
                xmax = as.Date("2021-01-01",  "%Y-%m-%d"),
                ymin = -Inf, 
                ymax = Inf),
            fill = "gray", 
            alpha = 0.009)
we <- we + annotate(geom="text",x=as.Date("2020-07-01"),
                                y=6.5,label="Summer 2020 surge", size = 3) + 
  annotate(geom="text",x=as.Date("2020-12-03"),
           y=6.5,label="Winter 2020 surge", size = 3)
we

Tie them all together

graphs <- ggpubr::ggarrange(Analytic,Cogproc,i,we,ncol=2, nrow=2, common.legend = TRUE, legend = "bottom")
annotate_figure(graphs,
                top = text_grob("CEOs' Language Change",  color = "black", face = "bold", size = 20),
                bottom = text_grob("Note. Vertical Line Represents the onset of the pandemic. \n\ Horizontal shading represents Standard Error. Vertical bars represent virus surges."
                                   , color = "Black",
                                   hjust = 1.1, x = 1, face = "italic", size = 16))

Package Citations

report::cite_packages()

##   - Grolemund G, Wickham H (2011). "Dates and Times Made Easy with lubridate." _Journal of Statistical Software_, *40*(3), 1-25. <https://www.jstatsoft.org/v40/i03/>.
##   - J L (2006). "Plotrix: a package in the red light district of R." _R-News_, *6*(4), 8-12.
##   - Kassambara A (2020). _ggpubr: 'ggplot2' Based Publication Ready Plots_. R package version 0.4.0, <https://CRAN.R-project.org/package=ggpubr>.
##   - Kuhn M (2022). _caret: Classification and Regression Training_. R package version 6.0-93, <https://CRAN.R-project.org/package=caret>.
##   - Lin G (2022). _reactable: Interactive Data Tables Based on 'React Table'_. R package version 0.3.0, <https://CRAN.R-project.org/package=reactable>.
##   - Müller K, Wickham H (2022). _tibble: Simple Data Frames_. R package version 3.1.8, <https://CRAN.R-project.org/package=tibble>.
##   - R Core Team (2022). _R: A Language and Environment for Statistical Computing_. R Foundation for Statistical Computing, Vienna, Austria. <https://www.R-project.org/>.
##   - Rinker TW, Kurkiewicz D (2018). _pacman: Package Management for R_. version 0.5.0, <http://github.com/trinker/pacman>.
##   - Robinson D, Hayes A, Couch S (2022). _broom: Convert Statistical Objects into Tidy Tibbles_. R package version 1.0.1, <https://CRAN.R-project.org/package=broom>.
##   - Sarkar D (2008). _Lattice: Multivariate Data Visualization with R_. Springer, New York. ISBN 978-0-387-75968-5, <http://lmdvr.r-forge.r-project.org>.
##   - Spinu V (2022). _timechange: Efficient Manipulation of Date-Times_. R package version 0.1.1, <https://CRAN.R-project.org/package=timechange>.
##   - Torchiano M (2020). _effsize: Efficient Effect Size Computation_. doi:10.5281/zenodo.1480624 <https://doi.org/10.5281/zenodo.1480624>, R package version 0.8.1, <https://CRAN.R-project.org/package=effsize>.
##   - Wickham H (2016). _ggplot2: Elegant Graphics for Data Analysis_. Springer-Verlag New York. ISBN 978-3-319-24277-4, <https://ggplot2.tidyverse.org>.
##   - Wickham H (2022). _forcats: Tools for Working with Categorical Variables (Factors)_. R package version 0.5.2, <https://CRAN.R-project.org/package=forcats>.
##   - Wickham H (2022). _stringr: Simple, Consistent Wrappers for Common String Operations_. R package version 1.4.1, <https://CRAN.R-project.org/package=stringr>.
##   - Wickham H, Averick M, Bryan J, Chang W, McGowan LD, François R, Grolemund G, Hayes A, Henry L, Hester J, Kuhn M, Pedersen TL, Miller E, Bache SM, Müller K, Ooms J, Robinson D, Seidel DP, Spinu V, Takahashi K, Vaughan D, Wilke C, Woo K, Yutani H (2019). "Welcome to the tidyverse." _Journal of Open Source Software_, *4*(43), 1686. doi:10.21105/joss.01686 <https://doi.org/10.21105/joss.01686>.
##   - Wickham H, François R, Henry L, Müller K (2022). _dplyr: A Grammar of Data Manipulation_. R package version 1.0.10, <https://CRAN.R-project.org/package=dplyr>.
##   - Wickham H, Girlich M (2022). _tidyr: Tidy Messy Data_. R package version 1.2.1, <https://CRAN.R-project.org/package=tidyr>.
##   - Wickham H, Henry L (2022). _purrr: Functional Programming Tools_. R package version 1.0.0, <https://CRAN.R-project.org/package=purrr>.
##   - Wickham H, Hester J, Bryan J (2022). _readr: Read Rectangular Text Data_. R package version 2.1.3, <https://CRAN.R-project.org/package=readr>.
##   - Zeileis A, Grothendieck G (2005). "zoo: S3 Infrastructure for Regular and Irregular Time Series." _Journal of Statistical Software_, *14*(6), 1-27. doi:10.18637/jss.v014.i06 <https://doi.org/10.18637/jss.v014.i06>.
##   - Zhu H (2021). _kableExtra: Construct Complex Table with 'kable' and Pipe Syntax_. R package version 1.3.4, <https://CRAN.R-project.org/package=kableExtra>.

All credit goes to the great Dani Cosme for teaching me how to make these! You can find her github here!